# Geared stepper maximum RPM

Hi, it might be that anyone knows here of what is the maximum gearbox end shaft RPM I could get with the geared stepper motor, for example this:http://www.phidgets.com/products.php?category=23&product_id=3325_0
The 904 RPM noted in the specs, is the output shaft RPM or the maximum that I can speed up the geared motor itself, thus giving me the shaft RPM in this case 904/5,18 = 174 RMP?

Do I understand correctly that the same motor without the gearbox.. http://www.phidgets.com/products.php?category=23&product_id=3324_0
..can make 4130 RPM, and the difference is because the gearbox reduces maximum speed of the motor?

Is this the link to the particular motor you are interested in? Your reference is to a whole page of stuff.

My reading of that page is that 904 is the RPM after the reduction gear.

You should keep in mind that if the motor is to rotate at 4000rpm or 67 revs per second it will need about 13,000 step pulses per second or one pulse every 76 microsecs. That might not leave an Arduino much time for anything else.

Robin2:
Your reference is to a whole page of stuff.

@arisetobe, if you use the "Quote feature on Robin2's post you can see how to add links properly.

If you don't want to hyperlink text as Robin2 did you could just add the URL tags like this:

If you quote this post, you'll see the tags I used.

It would probably be a good idea to let us know what you're trying to do. It sounds like you want to spin something fast. Is this correct? If so, how big and how heavy is the object to be spun and how fast do you want it to spin? Is it important to know the speed? How fast does it need to accelerate?

Speaking of acceleration, stepper motors generally can't be started at their full speed. They need to be accelerated which isn't very hard but neither is it a trivial task.

Does it need to be a stepper motor? There are lots of DC motors which will spin very fast without a gearbox and DC motors don't need an sort of special algorithm to accelerate. They do this on their own once power is applied.

If their controller can manage 4000rpm (which isn't impossible, but is pushing it), then the available
torque will be tiny at that speed - stepper torque drops off with speed

Hey guys, thanks for your quick feedback. I have just updated the links, sorry.

So, my story:

1. I am experimenting with 2 motors now, this: 28STH32 NEMA-11 Bipolar Gearless Stepper Motor - 3320_0 at Phidgets and this: 42STH38-1684B - 1.8 Degree - 1.68A Stepper - 5.18:1 Gearbox - 3325_0 at Phidgets both with the encoders so that I can read RPMs

2. I have Arduino UNO and Arduino DUE connected to a DRV8825 driver (with proper current limit according to the specs of the motors, 0.67A and 1.68A accordingly).

3. I am using AccelStepper library v1.5 (with proper microseconds speed update) as well as Encoder.h library to easily read encoder revolutions.

4. The Arduino UNO was fast enough to spin the NEMA11 motor up to 460 RPM (around max of it's specification), but it was the maximum for NEMA17 motor as well, thus divided to the gearbox rate I got 460/5.18 = 88RPM on end shaft of the geared 5.18 motor.

5. With Arduino DUE I could spin the NEMA17 geared motor up to 945RPM which is a similar number to 904RPM as in the specs (it is just I am not sure how to read the specs), but still divided to the rpm gear I get about 945/5.18= 182RPM.

6. With any combinations of accelerations and speed with the NEMA17 motor and the Arduino DUE did not give me the speed any higher than the 945 RPM. Seems that the DUE can spin faster (The UNO could not), but the motor just stalled, no matter how smoothly I am trying to accelerate it. The only last chance would be is to increase the current above the specs, which I don't want to do now.

7. I have also investigated that similar motors in NEMA17 size (for example these: ST4118 - Stepper motor – NEMA 17 | NANOTEC - select ST4118M1804 there, it has similar current and length) can be spinned to about 1000RPM, and if above, their torque falls dramatically, thus I have assumed that the behaviour of my motor is actually correct and this is the maximum I can get from the motor with the gearbox, as spinning above 1000RPM is not possible because of the gearbox resistance.

8. So, before selecting proper motors I need to clearly understand their capabilities and behaviour with gearbox, because for example this motor: 42STH38-1684B - 1.8 Degree - 1.68A Stepper - 27:1 Gearbox - 3327_0 at Phidgets it might be enough for me to have 164RPM (or even 100 RPM) from the gearbox, but it is not enough if I should divide it to the gearbox ratio, like this 164/26.85 = 6 RPM.

9. If my calculations are wrong and the specs are correct, there are some other questions then:
a. What such magical the controller, that Phidgets sells (PhidgetStepper Bipolar HC - 1067_0 at Phidgets), can do so it is better than the tiny DRV8825 and if I should invest buying it? Phidgets seem to use A4989 microchip, which seem to be below the DRV8825 in specs, but maybe Phidgets controller can accelerate the motor more smoothly, right? But the torque will fall dramatically then, and thus the speed will be useless..
b. Are there any alternatives to the 95\$ Phidgets controller which can be better than the DRV8825?

P.S. I am building a robotic arm. And I want to squeeze most from the motors, both speed and torque.

What voltage are you feeding to the stepper driver? Higher voltage should help with speed.

I seem to remember reading somewhere that the AccelStepper library is not good for high speeds. It may be worth experimenting without using any library. This Simple Stepper Code might be a starting point. But for very high pulse rates you would probably need to use port manipulation.

Clearly it would be unreasonable to expect a motor to start at very high speed but I suspect that, for testing, you would get away with a fairly crude steps-of-stairs type of “acceleration”.

However, as I said earlier, at very high step rates a 16MHz Arduino may have no time left for anything else.

…R

Robin2:
What voltage are you feeding to the stepper driver? Higher voltage should help with speed.

I am feeding the driver with 12v and 2.25A, while using the current limiter I get the voltage on the "ref" pin to be 0.84v, which, according to the current limiter explanation here Pololu - DRV8825 Stepper Motor Driver Carrier, High Current (md20a) (Current Limit = VREF × 2) gives me 1.68A according to the specs.

Robin2:
I seem to remember reading somewhere that the AccelStepper library is not good for high speeds. It may be worth experimenting without using any library. This Simple Stepper Code might be a starting point. But for very high pulse rates you would probably need to use port manipulation.

Clearly it would be unreasonable to expect a motor to start at very high speed but I suspect that, for testing, you would get away with a fairly crude steps-of-stairs type of "acceleration".

This is why I have found the library v1.5 which they say has improvements in speed by using microseconds. http://classes.engr.oregonstate.edu/engr/spring2011/engr421-001/group/gp8/MATLAB_CODE/Arduino%20AccelStepper/doc/ Running the motor without the library won't allow me to easily accelerate (unless you show me how do I do it) and I will have a stall right away.

Robin2:
However, as I said earlier, at very high step rates a 16MHz Arduino may have no time left for anything else.

I am using DUE and it is fast compared to UNO (which I tested as well), it seem not to be the board issue, but driver or library or motor.

I have no experience of the DUE but I guess it should solve any library speed problems.

Try powering the motor with 24v or more (subject to the upper limit allowed by you stepper driver). The higher voltage will ensure that the current reaches the limit more quickly.

...R

24V will get about twice the performance out of the motor, 48V, four times... Check what
the maximum voltage of your driver chip/board is first of course(!). Some microstepping may
reduce mid-band resonance which can stall a stepper well before its maximum speed.
Mechanical damping will also help with mid-band resonance and extra moment of inertia
in the load will change the freqency of the resonance.

An unloaded stepper on full steps is the most resonant configuration and is very prone to stalling/
mis-stepping and can behave differently sitting on a desk, held in the hand or clamped down...

Belt drives are quite good at damping resonance, leadscrews less so. I suspect the gearbox on
yours will help damp resonance nicely (if its low-backlash)

Thank you, the driver allows up to 45V, but the motor at least on Phidgets website says 12VDC is recommended.. Though on another website I see similar motor can be connected both to 12-24V.. http://www.omc-stepperonline.com/gear-ratio-51-planetary-gearbox-with-nema-17-stepper-motor-17hs130404spg5-p-140.html

Do you think I will burn the motor with the 24V?

Its a current controlled stepper motor, phrases like "12V recommended" are nonsense. The voltage
limit is the insulation breakdown strength, which will be more than 12V. You can assume all low impedance
bipolars are good to at least 48V I reckon. Many will run at 100V or more, since this is what gives
high performance for big CNC machines.

I suspect what Phidgets are getting at is it won't work well at 5V.

The key limit is the max safe rotation speed - ensure you don't go near this, as this means the
motor could sieze up or explode, but that's unlikely to be reachable in a small stepper.

Btw, I have ordered this AMIS-30543 driver Pololu - AMIS-30543 Stepper Motor Driver Carrier, must be better than 8825 and not that expensive as Phidgets.

An interesting device - especially if it can provide 2.5 amps or more. I had not seen that before.

I wonder if anyone else has used it.

I will also be interested to know if your assumption that it "must be better than 8825" is borne out in practice.

...R

yep, it is a new one.. This thing seem to be especially sweet: "SPI-programmable current control (from 132 mA to 3 A) enables your microcontroller to adjust the peak-current limit on the fly as more or less torque or speed is needed". I will let you know how it works.

So, I am getting the idea about the voltage increase for steppers from here: Stepper Motors and Power Supplies · synthetos/TinyG Wiki · GitHub and would like to order a universal 12v-24v 5A power supply, just for experiments..

Now, MarkT, could you please clarify this: "24V will get about twice the performance out of the motor, 48V, four times...". Sorry if my question is a bit "nooby", but if we are talking about current that drives motor, and there is current limiter on the driver, what exactly the switching from 12 to 24V will provide? Torque? Speed?

And what I will need is just to re-adjust the current limiter on the driver and the only thing I should worry about is that wires do not overheat?

arisetobe:
and would like to order a universal 12v-24v 5A power supply, just for experiments..

I seem to remember reading that power supplies which control their output currents don't work well with stepper drivers - both devices fight with each other while trying to control the current. Check this yourself before spending money. I may be wrong.

could you please clarify this: "24V will get about twice the performance out of the motor, 48V, four times...". Sorry if my question is a bit "nooby", but if we are talking about current that drives motor, and there is current limiter on the driver, what exactly the switching from 12 to 24V will provide? Torque? Speed?

As I mentioned in Reply #7 the higher voltage causes the current limit to be reached more quickly so that a greater proportion of each step interval is spent at full current. The inductance of the coils and the back emf generated by the moving motor oppose the build up of current.

And what I will need is just to re-adjust the current limiter on the driver and the only thing I should worry about is that wires do not overheat?

It is the wires in the coils inside the motor that are at risk of overheating and damage.

...R

Thank you Robin2 for the explanation and suggestions. I do not get however the phrase "power supplies which control their output currents ", are there any that do not do that? Is there a way you can distinguish them?

arisetobe:
I do not get however the phrase "power supplies which control their output currents "

I am not an expert on this but I believe you can get power supplies for your test bench that allow you to adjust both the output voltage and the maximum output current (perhaps not both at the same time?)

You can also get power supplies with over-current protection but that should not be a problem as long as the maximum current is well in excess of what your motor requires.

...R

I have just read about the differences of regulated and unregulated power supplies. It also led me to conclusion that:

1. Most of currently commercially available power supplies are regulated.
2. Batteries are unregulated power supply.

Does this mean that I might get issues with steppers if I use regulated power supply and there will be likely no issues if I use batteries?

Batteries are not suitable because stepper motors are very inefficient. They need full power even when stationary.

Just get a power supply that can provide at least twice as much current as you need at the voltage you require.

I understand from another Forum contributor that an unregulated supply with a big smoothing capacitor is best. But I have my (low current) motors working off a 19v laptop power supply.

All I was trying to warn you against was a power supply where you can set the maximum current and where you may be tempted to set the power supply current at a level that protects the motor. Leave it to the stepper driver to do that.

...R

Ah, I see, ok, got it, thanks.

Regarding batteries, after I complete the robotic arm I will continue to create a biped.. and I will need the batteries eventually. I have found some.. Custom NiMH Battery Pack: 24V 10 Ah with Anderson Connector seem to be ok for the scale of my project.. But that is another story.. I will write here when I test the motor with 24V.. curious if it makes the motor stronger and faster.